Tanishq Gautam

Computer Vision Engineer 🏆 @Kaggle Notebooks Master (Top 75) + 3X Expert 📈 @weights_biases Dev Expert 🔥 @Pytorch Aficionado

[Tanishq Gautam @Closedcv]

I finally made it into the top 100 Kagglers in Notebooks :')

[Aahan Singh @aahan_builds]

Day 8 done. The dashboard is live.

https://t.co/6GS4cNDeXT

Built with Astro SSR + Cloudflare Pages.
Bloomberg-dark aesthetic.
Mock data today. Live pipeline next.

The alerts feed is the page I'm most proud of.
Every divergence signal. In order. With the reasoning.

$TSLA DOUBLE OVERBOUGHT +91 | RSI 76.8
$NVDA WHALE EXIT +87 | Druckenmiller 13F full exit
$GOOGL SENTIMENT SHIFT | -42pts in 2 hours
6 days of backend work.
This is what it looks like.

[Aahan Singh @aahan_builds]

Day 3/14 of @CrowdVsWhale 🐳. PR merged.

Hardest day yet.
90% spent learning Agno and building a readable pipeline.

CROW now runs as a proper Agno workflow step. Not a script. A real agent in a real pipeline.

Today's result: ran CROW on $CBRS Cerebras analysing @cerebras. Screenshot below.

@CrowdVsWhale built on @AgnoAgi run on @cerebras 🔥🔥🔥🔥🔥

Github -> https://t.co/KjNkyR18bu

#buildinpublic #AI #agents #fintech

[Aahan Singh @aahan_builds]

I'm building CrowdVsWhale in 14 days. Day 1 done. 🐳

Thread on what I shipped today and what this thing actually is 👇

[Tanishq Gautam @Closedcv]

@AahanSingh @hashnode

[Tanishq Gautam @Closedcv]

@SinghMadhu302 Thank you!

[Tanishq Gautam @Closedcv]

Look at me now, Mom and Dad! I'm an #MLOpsGenius

Thanks! #Valohai

[Tanishq Gautam @Closedcv]

@Pr1yankagautam Thank you ❤️

[Stella Biderman @BlancheMinerva]

So… apparently @MetaAI is going to be releasing a 175B parameter language model for anyone to download

https://t.co/56nYUpmsTn

[Matthew Carrigan @carrigmat]

Saw this in the @huggingface office, went back to working on my laptop, and then four hours later shouted "OH I GET IT, BECAUSE IT'S A FORK OF CAFFE"

[Mark Tenenholtz @marktenenholtz]

I spent 400+ hours training computer vision models last year on my road to Kaggle Master.

I recently revisited my code and notes from those competitions and distilled them into a repeatable process that anyone can follow.

7 steps to train any computer vision model 🧵

[Andrej Karpathy @karpathy]

Computer vision research feels a bit stagnating in a local minimum of 2D texture recognition on ImageNet, COCO etc. This is great but only step 1. Unlocking further progress needs new framework:
1) the data source has to become diverse videos, not individual frames from internet

[Mark Conway @_MarkConway_]

After 8 years and 137 competitions, it’s time to retire from @kaggle. Highest rank was #183. If only they awarded medals for March Madness back then… 🥇

[Bojan Tunguz @tunguz]

More incredible work by OpenAI. They have built a neural theorem prover that can successfully solve many high-school Math Olympiad problems. The prover uses a language model to find proofs of formal statements.
https://t.co/drVQQFWrgu

#NLP #ML #DL #AI

[Yujia Li @liyuajia]

Excited to share the project #AlphaCode I’ve been working on for more than 2 years! Can’t believe we started before COVID is a thing and worked through this project mostly at home, with an amazing team!

[Data Science Dojo @DataScienceDojo]

AI Best: 70+ Machine Learning Algorithms in Python: https://t.co/6i8FZHZt77

[Tanishq Gautam @Closedcv]

Highly recommend checking out this article that expands on end to end design case studies!

[ohshuxbro @ohshuxbro]

Finally!!
Published my first #uxdesign article with @uxdesigncc on Bootcamp! So pumped!! #uxdesign #designthinking #designsprint #UserExperience #uxresearch #uiux
https://t.co/tKxfuiKevp

[Bojan Tunguz @tunguz]

Can confirm that taking a break from @Kaggle can do miracles for your mental, spiritual, and physical health. 10/10, highly recommended.

[Tanishq Gautam @Closedcv]

I know I'm late but I finally got access to Github Copilot! 🤖

[Tanishq Gautam @Closedcv]

Things we love to see!

[Weights & Biases @wandb]

We're thrilled to announce that we've raised a $100m+ Series C to continue building our developer-first MLOps platform.

We'd like to thank our users, community & friends – this would not have been possible with you.

https://t.co/o32vbd6Qne

[Andrej Karpathy @karpathy]

Errr ok wow, I am shook by the new ConvMixer architecture
https://t.co/crUMktQ0ig "the first model that achieves the elusive dual goals of 80%+ ImageNet top-1 accuracy while also fitting into a tweet" 😐